Industrial cameras bring vision to modern eye tracking technologies

Since its beginnings in the late 1870’s, eye-tracking has gradually evolved and modern systems now rely on machine vision cameras and sophisticated software algorithms to track eye movements.

LC Technologies, Inc.(Fairfax, Virginia, USA) is a leading provider of flexible and open eye tracking technology with over 25 years’ experience in designing a wide variety of systems used in assistive technologies, marketing and medical research, and more.

Eye Tracking Technology Empowers People with Disabilities
LC Technologies’ Eyegaze Edge® is an assistive technology system that enables people with disabilities such as ALS, Cerebral Palsy, Muscular Dystrophy, Multiple Sclerosis, spinal cord injury, stroke (Locked-in Syndrome) and traumatic brain injury to stay connected with the world. Eyegaze Edge allows users, who are unable to control their body movement or speak, to speak, write, operate a computer, play video games, control appliances and more simply by using their eyes.

Eyegaze Edge is a non-intrusive monocular system based on a single machine vision camera that is used to track one eye of the user to determine the gaze point in order to operate the system. It consists of a 13.3” tablet PC, a Guppy F-038 NIR industrial camera by Allied Vision Technologies mounted just below the screen, and an adjustable table mount.

The Guppy F-038 NIR camera is a very small camera featuring an interlaced CCD sensor with enhanced near-infrared sensitivity that makes it ideal for imaging in low-light conditions. The camera system features LC Technologies’ patent aperture technology, which incorporates a small, low-power infrared LED light at the centre of the camera lens. The infrared light is used to illuminate the eye to generate a corneal reflection that causes a “bright pupil effect”. The camera frames the user’s pupil and images at 60 frames per second to capture the user’s eye movements.

Sophisticated image processing software continually analyzes the video image of the eye in real-time and determines where the user is looking on the screen thanks to trigonometric calculations based on the position of the user’s pupil centre and corneal reflection within the image. Gaze point measurements are made with ¼ inch accuracy or less.

Users operate the Eyegaze Edge by looking at rectangular keys that are displayed on the control screen and are able to “press” a key by looking at it for a specified period of time. The system features an array of menus and options that allow them to navigate the system and perform a multitude of tasks from synthesizing speech, controlling their environment (lights, call bells, etc.), typing, running computer software, operating a computer mouse, accessing email and more.

A 15 second calibration procedure is required to set up the system. To calibrate, the user fixes his/her gaze on a sequence of small circles displayed on Eyegaze Edge‘s screen. This allows the system to learn several physiological properties of the person’s eye, such as the curvature of the eye’s cornea, in order to be able to project his/her gaze point accurately. The system can be adjusted to accommodate users with very large and very small pupils, drooping eyelids, and irregularly shaped pupils. It can also be operated in any position (sitting, lying down or head tipped to the side).

EyeFollower for Research LC Technologies’ EyeFollower is a binocular system that features two Guppy F-038 NIR machine vision cameras mounted on a gimbal to allow subjects to move freely while capturing gaze points data within a designated area . The EyeFollower is able to track and locate users as well as adjust and re-calibrate itself with 0.4 degree accuracy thanks to image processing algorithms. The system uses patented asymmetric aperture method to measure variations in the range between the camera and the cornea of the eye as well as range information to minimize gaze point tracking errors resulting from longitudinal head motions.

The EyeFollower is employed in a multitude of applications such as marketing, medical research, and pilot training. For example, in the case of a high investment print advert, eye tracking systems are able to provide advertisers with a full report of which part of the advert people looked at and for how long to determine whether the current advert design is effective or not. The provided metrics (first fixation, most fixations, gaze time, fixation order and movement between fixations) are recorded and presented through visualization tools such as heat maps or gaze plots. Medical applications include research and the diagnosis of psychological illnesses or neurological trauma following a brain injury. By analyzing a patient’s gaze doctors are able to make a diagnosis based on gaze patterns and behaviours. Flight simulators also use eye tracking technology to evaluate the progress of pilots in training by comparing gaze points patterns between experienced and novice pilots.

What’s next for Eye Tracking?
As eye tracking technology continues to evolve, one of the last hurdles is to overcome cost and portability, which would enable the technology to be implemented in any person’s daily routine.

Eye tracking, for example, could be implemented in the automobile industry with in-car interaction systems allowing drivers to turn on their radio or adjust temperature settings by simply looking at a control on the car’s windshield and without having to look away from the road or move their hands from the steering wheel. Other car safety applications would include the car knowing whether you are paying attention to the road or not based on where your eyes are looking. In-home applications include being able to control home appliances (TV, light switch, heat thermostat) just by looking at a control. Within the video game community our eyes can be used in virtual reality type applications where elements in a game, such as looking around the environment, can be controlled by where the subject is looking.

For more information, please visit: www.alliedvisiontec.com

• CID: 14517